Medicare's 2026 dementia data push focuses on building better national surveillance, not on measuring any one person's cognitive future. HHS is updating its National Plan to Address Alzheimer's Disease through 2035, and the agency's own research shows that claims-based dementia counts and survey-based counts diverge so widely that a "millions undiagnosed" headline describes a population-level measurement gap, not a forecast about someone already diagnosed with mild cognitive impairment (MCI).
For a reader with MCI, the useful question isn't whether the headline is alarming. It's whether the underlying data problem has anything to do with their own diagnosis, their own doctor, or their own records. It mostly doesn't, and understanding why makes the headline easier to read without overreacting to it.
Medical information disclaimer: This article is for general educational purposes only and does not provide medical advice, diagnosis, or treatment. Always consult a physician or other qualified health professional about symptoms, medications, tests, or treatment decisions.
Official resources:
- Check eligibility on NIH’s official study page — Use this page to check the study’s participation requirements.
- Read the original research from NIH — Use this page to review the original research.
Table of Contents
- What is HHS actually updating in 2026?
- Why do Medicare claims undercount dementia and MCI?
- Is dementia undercounted, or is it specifically MCI?
- What actually determines whether MCI shows up in your Medicare record?
- How should someone with an MCI diagnosis read these headlines?
- Frequently Asked Questions
What is HHS actually updating in 2026?
On July 24, 2026, HHS's Assistant Secretary for Planning and Evaluation (ASPE) published a Request for Information in the Federal Register to inform a comprehensive update to the National Plan to Address Alzheimer's Disease, extending the plan's goals through 2035. The RFI asked for input on research, care, public-health, and data-infrastructure priorities, and public comments closed August 15, 2026, according to the Federal Register notice. Alongside that planning process, a CDC-authored federal agency update describes the agency deploying roughly $41.5 million in fiscal year 2026 for the BOLD Alzheimer's public-health infrastructure program — its largest annual investment in that program to date.
The same update describes CDC expanding its Behavioral Risk Factor Surveillance System (BRFSS) modules: a subjective cognitive decline module now running in 32 states and a caregiving module running in 27 states, per the CDC/HHS federal agency update. Both efforts are about counting people and tracking trends across states, not about changing how any individual's diagnosis is made or managed. The data-infrastructure language in the RFI and the BRFSS expansion both point toward the same goal: closing known gaps in how dementia and cognitive decline are measured nationally.
Why do Medicare claims undercount dementia and MCI?
Medicare claims data — the billing codes doctors submit for every visit — is one of the main sources used to estimate how many Americans have dementia or MCI. But claims only capture a diagnosis if a clinician codes it, and a 2026 study in the Journal of the American Geriatrics Society found that claims-based dementia algorithms catch far fewer true cases than survey-based assessment does. The study, led by Gupta and colleagues, compared Medicare claims algorithms against the National Health and Aging Trends Study (NHATS), a survey that classifies participants as having "probable" dementia through direct cognitive testing and informant interviews. The claims algorithms correctly identified only 48.8% to 70.5% of people NHATS classified as having probable dementia — their sensitivity — even though the same algorithms correctly ruled out people without dementia 91.5% to 97.8% of the time, according to the JAGS study.
In plain terms: claims data is good at confirming who doesn't have dementia, but it misses a substantial share of who does. A related validation study extended this comparison across Medicare's two main coverage types. It found specificity above 97.9% and negative predictive value above 92.9% in both Medicare fee-for-service and Medicare Advantage, but sensitivity that was lower, or no better, in Medicare Advantage than in traditional fee-for-service, according to the claims-validation study. That matters because which type of Medicare coverage someone has can affect how completely their cognitive diagnosis is reflected in the claims record used for national estimates.
Is dementia undercounted, or is it specifically MCI?
The two conditions are not undercounted by the same margin. An observational analysis using 2017-2019 Medicare data estimated about 7.3 million beneficiaries had undiagnosed MCI in their claims record, compared with only about 484,000 beneficiaries with undiagnosed dementia, according to the Neurology observational study. MCI is the condition driving most of the measurement gap, not dementia as a whole. That gap traces partly back to how NHATS itself is built.
NHATS classifies people into "probable," "possible," or "no" dementia categories using self- or proxy-reported doctor diagnoses, an informant AD8 cognitive screening score, and direct cognitive tests, according to the NHATS/NSOC methodology documentation. It does not include a distinct clinical MCI category. So the main survey used to validate Medicare's dementia algorithms isn't designed to separately track MCI in the first place, which helps explain why MCI's claims-based undercount is so much larger than dementia's. For someone already diagnosed with MCI, this distinction is the most important fact in the whole data story. The "millions undiagnosed" figure describes people whose MCI has never been coded in Medicare claims at all — a measurement category that, by definition, doesn't include someone who already has a documented diagnosis.
What actually determines whether MCI shows up in your Medicare record?
Whether a person's MCI gets coded in Medicare claims depends heavily on the type of visit they use, not on the severity of their symptoms. A University of Texas Medical Branch-linked analysis found that Medicare Annual Wellness Visits were associated with a 21% increase in documented MCI diagnoses compared with other visit types, according to the UTMB Sealy Center on Aging analysis.
Medicare's Annual Wellness Visit includes a required cognitive assessment that a routine problem-focused visit often doesn't. This points to a practical, concrete takeaway that has nothing to do with the headline's national numbers: using the Annual Wellness Visit, rather than relying only on symptom-driven appointments, is the visit type most likely to generate a documented cognitive assessment in the first place. For someone managing MCI, scheduling that visit — and confirming the cognitive assessment is actually performed and coded — is a more useful action than reacting to a population-level undercount statistic.
How should someone with an MCI diagnosis read these headlines?
A story reporting that millions of Americans have undiagnosed dementia or MCI is describing a surveillance estimate built from comparing claims data to survey data across a national sample. It is not describing any individual's risk of being missed, misdiagnosed, or reclassified. The sensitivity gaps measured in the JAGS study and the 2017-2019 Medicare analysis above are population statistics, and reading them as a personal prediction about diagnosis or progression isn't what the data supports, according to the synthesis of these findings.
The practical distinction to hold onto: these numbers measure how many people's conditions never got coded at all. Someone who already has an MCI diagnosis on record has, by definition, cleared the exact hurdle these studies say millions of others haven't. The headline is a public-health measurement story — useful for shaping where CDC and HHS direct surveillance dollars and survey coverage, as reflected in the BOLD program funding and BRFSS expansion — not a signal about any one diagnosis already in hand.
Frequently Asked Questions
Does the "millions undiagnosed" MCI estimate include people who already have a documented diagnosis?
No. The 7.3 million figure from the 2017-2019 Medicare analysis counts only beneficiaries whose MCI never appeared in claims data at all, according to the Neurology observational study.
Why is MCI undercounted so much more than dementia in Medicare claims?
The main survey used to validate these claims algorithms, NHATS, sorts people into dementia categories but has no separate clinical MCI classification, according to the NHATS/NSOC methodology chapter, which limits how precisely MCI specifically can be cross-checked against claims.
Does the visit type matter for getting MCI documented?
Yes. Medicare Annual Wellness Visits, which include a required cognitive assessment, were linked to a 21% increase in documented MCI diagnoses compared with other visit types, according to the UTMB Sealy Center on Aging analysis.
You Might Also Like
- Healthy Diets in the 2026 WHO Dementia Guideline: What People With Mild Cognitive Impairment Can Track Without Self-Diagnosing
- Food Access and Brain Health Research Priorities in 2026: A Source-Checked Research Briefing for People With Mild Cognitive Impairment
- Healthy Diets in the 2026 WHO Dementia Guideline: Access and Adherence Barriers for People With Mild Cognitive Impairment
